Mine Fan Intelligent Faults Diagnosis Based on the Lifting Wavelet Packet and RBF Neural Network

In order to overcome the disadvantage of traditional methods of fault features extraction, and realize the online and intelligent fault diagnosis, a new method of feature extraction based on the lifting wavelet packet transform was presented, with which fault feature factors were extracted from three typical running states of mine fan. The fault feature factors can be taken as the input samples of RBF neural network, which realized the intelligent fault diagnosis of mine fan. The results showed that the combinative method of the lifting wavelet packet decomposition and RBF neural network can reduce the need of time and memory greatly, and it is very fit for the real-time and intelligent conditions monitoring and fault diagnosis of machinery system.